ELECTRONIC TIMERS - AN OVERVIEW
By Chris Edge and Martin Gregorie
Electronic timers have been used in free flight models since the late 1970's. Recent developments in micro-chip technology and innovative use of R/C servo drives has resulted in a greater range of timers being available than ever before. This paper aims to provide an overview of current systems with their advantages or otherwise and hint at future developments.
HISTORY
Individuals have tried to develop electronic timers since the mid 1970's. One of the key drivers in the field has been Thomas Køster who demonstrated integrated systems for F1C models in 1977. Early systems soon proved that the F1C environment, with its high frequency vibration and fuel in liquid and vaporised forms, prevented the use of simple electromechanical switching to set the function times. Thomas solved this problem by introducing an external control box to program the settings into an onboard memory chip, and using a solid state Hall effect start switch. This approach entirely eliminated mechanical switches from the model. The timer output was via an R/C servo that released arms rather like a clockwork timer. The Køster digital F1C timer and a simple F1A timer became commercially available in 1982.
Despite continued use in the hands of Køster, the F1C timer never sold due perhaps to its complexity and certainly its cost. On the other hand the Køster F1A system, a single function fully self-contained unit, sold well and became the standard electronic timer until it ceased production in the late 1980's.
The advent of the bunt launch on F1A's and the need for multi-function timers has spurred development of electronic glider timers to the point where there are least four systems on the market and others in prototype form. In F1A at least these systems will doubtless become common.
Martin Gregorie and Chris Edge have been using electronic timers for F1A since their inception. The aim of this article is to comment on the advantages and disadvantages of the new systems.
SINGLE FUNCTION TIMERS
Our experience has shown that single function electronic timers show considerable advantages over clockwork devices but they have their own difficulties that need to be addressed.
Key advantages are accurate, unchanging timings up to (typically) a maximum of 9 minutes 54 seconds and the ability to provide a fail-safe D/T if the line breaks or is accidentally released. These advantages are only realised if the whole system is carefully maintained, rather like a clockwork timer. The difference in this case is that maintenance is more related to battery state, wire/connector integrity and switch integrity.
Batteries are usually multi-celled NiCds and can be charged continually at 1% of their maximum capacity. In this mode a charged battery can run for over 300 cycles. They are limited in their life but various monitoring methods have been developed (e.g. voltage drop under load) and in the worst case batteries are easy to replace.
The switches supplied with many timers have proved to be unreliable and we now recommend the use of sealed types from suppliers such as Honeywell. Connectors and wiring need to be checked if the timer is removed from the model or disturbed by a hard landing but never fail in flight.
The weight of a single function timer system is more than the equivalent clockwork timer; however the batteries can be placed in the nose as ballast. The other parts weigh little more than a KSB or Seelig; the timer electronics are approximately the same size as a mechanical timer.
MULTI-FUNCTION TIMERS
A twin-function timer was developed by Van Wallene for rudder delay and D/T but the more recent timers can operate a full bunt system in an F1A or supply all the timing functions for F1B and F1C. A number of different systems are available but common features are twin microswitches (for F1A), a minimum of three functions and flexibility in timings. The systems use a servo drive or a geared motor as the output device. It is the former which now appears to offer the greatest benefit but is equally the most expensive.
SERVO DRIVE TIMERS
Køster's original commercial power timer used a servo drive actuation and was set via an external control box. M&K, Bauer, Nyhegn/Køster, Gewain and others are now selling later versions type of timer. These provide the possibility of additional functions (over 10 in some cases) compared to clockwork systems. The basic concept is to drive the servo to pre-set positions at the programmed intervals after launch. The output from the servo can be a simple half disc that releases arms or a direct connection to the tailplane. F1C installations have tended to use the first approach while F1A systems have used the latter. Recent experience has shown that higher bunt launches are indeed possible with this increased timer flexibility.
Bauer has previously marketed an analogue servo drive system that relies on potentiometers to set servo positions and time intervals. Whilst this has been shown to work the use of potentiometers makes settings difficult to calculate and they tend to be vulnerable to the European climate!
The same advantages that single function electronic timers have over mechanical timers apply to multi-function timers but there are some additional benefits. Short bunt times can now be changed by 'dialling them in' without fear of error. Furthermore the microswitch logic can be harnessed to sense when the flyer has failed to launch properly and is in danger of bunting into the ground. In such circumstances the timer will either D/T the model or revert to a non-bunt launch. This latter concept is being taken one stage further by Matt Gewain. He is now using a servo to drive the rudder as well thus allowing the electronics to re-configure the model from a bunter to a zoomer in the right circumstances.
Whilst commercial timers are now available the advent of cheap, readily available self-contained microcontrollers with built-in basic interpreters allows competent individuals to develop custom systems at a relatively low cost. One such chip, known as the Parallax Basic STAMP, has been used by Gewain in his system and by others in their ongoing developments.
MOTOR-DRIVE TIMERS
Maarten van Dijk and Allard Van Wallene have developed two systems that use simple motor drives rather than a servo to control a bunt F1A. The systems are limited in that they can only release arms; they can't provide the ability to directly control the tailplane.
Wallene's first system, perhaps the first programmable bunt system to be commercially available, is programmed via a separate control box. It has up to ten functions available. Problems with damage to the chip sets has led to the development of a simpler system which dispenses with the external control box. This utilises an electric motor and gearbox to release arms in a sequence controlled by a slotted disk. Whilst not having the flexibility of other systems its simplicity makes it a cheap and reliable unit.
PROBLEMS
The use of any electronics implies that a suitable battery must be used to drive both the timer and its output device. The electronics of single-function timers have extremely low power requirements. The D/T release mechanism, whether motor or solenoid, also requires little power and over 300 flights are possible on a single battery charge. Multi-function timers have much higher power requirements as both the microcontroller and the servo require considerably more power than their single-function equivalents. This in turn requires the timer designer to adopt a well thought-out power management strategy or the operational life between charges may become unacceptably short.
The fact that commercial systems are available and in the hands of the "non developers" means the problem isn't too great, although charging overnight after each day's flying is probably a prerequisite to avoiding a flat battery. Nickel hydride rechargeable batteries have greater capacity and are used by Nyhegn. However, these require a careful charging regime to get best performance.
The requirement to use two microswitches to fully monitor the hook state also has consequences. Whilst a microswitch in front of the hook is easy to install, as in the case of single- function timers, the requirement for a second switch to sense the position of the latch means that the hook now is literally wired into the model. Switch size is also critical but suppliers such as RS sell sub-miniature waterproof switches that have been used successfully. It is worth noting that gold-contacted, fully-sealed switches are the preferred solution for European conditions of high humidity and moisture.
FURTHER IDEAS
With the use of on-board memory and processing power a number of possibilities not available with clockwork systems present themselves.
Mention has already been made of the Gewain twin-servo system. A servo operates the tailplane as previously described but a second servo operates the rudder, using the hook microswitch logic to determine rudder deflection. The advantages are that the model can re-configure itself in cases where a bunt would be otherwise aborted to provide a zoom launch. It is not clear if the disadvantages of power consumption and installation complexity, amongst others, will allow this system to be commonly used. However it does indicate the potential for such systems.
Bauer has taken another approach to the premature unlatch problem by developing a hook than can re-latch. If the system detects that the hook has remained unlatched without the model being launched within a preset time the latch can be closed by a solenoid to re-lock the tow-line in place. Bauer has also discussed doing away with the traditional hook spring and utilising strain gauges fitted to the hook shaft to measure the tow tension and hence allow the microcontroller to determine the point at which the hook should unlatch. As far as is known he has not yet built a prototype of this system.
So far only passive systems have been discussed. There is no reason why on-board measurement of flight parameters can't be used to provide an optimum flight path for any given set of conditions. Gewain, for example, has developed an active sensor to dynamically control the bunt transition of an F1A. The microcontroller monitors the state of a mercury tilt switch so as to effect the start and finish of the bunt. So far the properties of the mercury switch limit its use to the non-vertical (soft) bunts used by Matt, however, Schlosberg (Reference 1) has suggested that using several mercury switches could offer scope for improvements to the system.
Another approach would be to measure the local angle of incidence or flight speed and use this data to calculate the moment of push over into the bunt and as well as the optimum point for recovery to gliding flight. Sensors could be used to activate rudder when rising air is detected or to detect when the model is flying over trees in order to extend D/T times. To some extent the future of on-board electronics is limited only by the mind of the imagine.
REFERENCES
1. Schlosberg A. - "Nordic Electronic timers and Mercury Switches" NFFS Symposium Report 1996
FOOTNOTE
In August 1997 Viktor Stamov won the F1A World Championship in Sazena using his model CB-62 equipped with the M&K electronic timer
